Hss4303b – Intro to Epidemiology March22, 2010 – Environmental Epidemiology
But first…. Today: class evaluation Thursday: Erin will teach the class Monday: guest lecture by Dr Nick Barrowman on sample size calculations
Environmental Epidemiology
the branch of public health that deals with environmental conditions and hazards that may pose a risk to human health. Environmental epidemiology identifies and quantifies exposures to environmental contaminants; conducts risk assessments and risk communication; provides medical evaluation and surveillance for adverse health effects; and provides health-based guidance on levels of exposure to such contaminants. – -Wikipedia
Environmental Epidemiology Exposures – Chemical spills – Pollution – Power lines – Heavy metals – etc Outcomes – Cancer – Reproductive limitations – Congenital malformations – mortality
Study Designs Descriptive – Setting priorities – Identifying hazards – Formulating hypotheses Analytic – Collecting evidence for causality
Study Designs Particular reliance on ecological designs – What is that again? e.g, correlate average mortality within a census tract with average exposure within a census tract
Study Designs Case-control are particularly popular – Why?
Occupational Studies Many (if not most) env epi studies are occupational Heavy confounding by “healthy worker effect” – What is it? 1.Since people who work are the healthiest, any effects will be lessened. 2.Non-malignant effects more easily detected, since malignant diseases may not manifest till later in life
Two concepts Latency Synergism
Two concepts Latency – Time between initial exposure and first onset of a measurable effect – Can vary from seconds to decades Synergism – Combined effect of several exposures is greater than the sum of those exposures Eg, asbestos increases risk of lung cancer by X % Smoking increases risk of lung cancer by Y % (smoking + asbestos) increases risk by >(X+Y)%
Example of an Env Epi study Ecological Case control (Some of the following slides from study by Mark Goldberg, McGill University, 2000)
In both studies… Exposure = proximity to landfill site – Ecological: crude “exposure zones” – C-c: distance from each person’s house – Zones defined as “high”, “high A” and “high B” Exposure defined at time of diagnosis (via interview) – What kinds of biases does this introduce? Recall bias Selection bias (selects for people living in area a long time, since cancers have long latency)
What’s a logistic regression?
To read more about this example Go to: –
Next time Before Monday, browse the two articles on the website on sample size calculation – Cohen (1992) – Donner (1984)